Much teaching about artificial intelligence begins from an assumption we want to contest: that educators and students have no option but to adapt to whatever systems happen to arrive on the market. This perspective paper treats refusal and redesign as legitimate pedagogical orientations. By refusal, we do not mean a generalized hostility to computation; rather, a principled rejection of uncritical or institutionally mandated adoption, especially when AI systems depend on surveillance or extractive data collection. Redesign is its constructive counterpart: the work of remaking a tool, a task, or an assessment so that it serves pedagogical rather than commercial ends. The centre of gravity is classroom practice. We argue that students need foundational critical AI literacy before they meet discipline-specific applications, and that this literacy is built through dialogic teaching in which ethical tensions are debated and held open. We outline a classroom sequence and assessment approach that makes refusal and redesign assessable: structured deliberation about when a system should be used, constrained, or declined, and assignments asking students to rebuild a flawed tool or assessment against criteria that reward reasoned judgement over compliance. What we hope to offer is a defensible shift in orientation, away from adaptation and toward critical agency. The horizon of that shift is collective. A classroom where declining a harmful system counts as a teachable, gradable competence is also a classroom where redesign opens onto more accountable technological futures. Students’ agency must be at the centre of critically informed AI adoption practices.
Dr. Johanathan Woodworth is an Assistant Professor in the Faculty of Education. His research focuses on educational technology, AI literacy, assessment validity, language teacher education, and the ethical use of generative AI in teaching and learning. He develops frameworks that support informed and defensible educational decisions about AI, including BEARA and the Validity Architecture for AI-Integrated Assessment. His scholarship examines the intersections of pedagogy, validity, critical AI literacy, equity, teacher judgment, and AI-mediated student work.
Dr. Cornelia Schneider hails originally from Germany where she did a special education degree at the Universität Würzburg, but then advanced to researching inclusive education at the University of Paris Descartes (Sorbonne) where she completed her Master and PhD degrees. Her research and teaching focuses on the perspective of the Sociology of Childhood and Disability Studies, recognizing children’s and disabled people’s agency. Cornelia utilizes a comparative approach to foster educational and inclusive frameworks, policies, and practices. She has been involved in research and policy work about inclusive education in Canada and other countries including France, Germany, Portugal, Hungary, Sweden and Honduras.
Dr. Andrea Fraser is an Assistant Professor in the Faculty of Education at Mount Saint Vincent University. She holds a PhD and MEd in Curriculum and Instruction, a post-degree certificate in Special Education, and a Bachelor of Education. Her research examines reading acquisition and development, universal screening, literacy pedagogy, and teacher knowledge and beliefs across pre-service and in-service contexts. She is particularly interested in how instructional and assessment practices are implemented, supported, and adopted by educators in support of student development. Her work also explores opportunities for supporting diverse learners and enhancing teacher preparation to promote equitable and effective literacy instruction.